dwishank/items_raw_full

该数据集包含800,000个训练样本、10,000个验证样本和10,000个测试样本，每个样本包括标题（title）、类别（category）、价格（price）、完整描述（full）、重量（weight）等字段；summary、prompt、id字段可能为空。The dataset contains 800,000 training examples, 10,000 validation examples, and 10,000 test examples. Each example includes fields such as title, category, price, full description, and weight; the summary, prompt, and id fields may be null.